With nearly a fifth of the population living with HIV, Botswana has one of the highest HIV prevalences in the world. Low-cost, scalable interventions are essential to support people living with HIV (PLWH) to adhere to antiretroviral treatment (ART) and remain in care. One such intervention is the use of treatment/adherence partners, which are recommended by HIV treatment guidelines in at least 20 countries world-wide. Specifically, to support adherence and reduce treatment discontinuation, national HIV policies of several countries, including Botswana, recommend that healthcare providers encourage patients initiating ART to identify an individual who can provide support, accompany patients to appointments, and provide reminders for medication. Treatment partners necessarily leverage patients' social networks, requiring that PLWH select an individual in their social circle to provie social support. Although a large body of work indicates the key role of social support in promoting adherence, research on the effectiveness of treatment partners has shown mixed results. Thus, research is needed to determine the ways in which support from treatment partners can be better harnessed for optimal effects on health outcomes. The proposed research objective is to determine the most effective characteristics of treatment partners from a social network perspective, in order to inform healthcare providers on how to guide PLWH on treatment partner selection. The specific aims are: 1) To conduct a social network analysis of people living with HIV and their treatment partners in Botswana in order to compare retrospectively individual-, dyadic-, and social network-level characteristics that are significanty associated with virologic failure; 2) To qualitatively explore factors related to treatment partner selection, including perceived barriers to and facilitators of selection of effective treatment partners; and 3) To use study results to develop and disseminate messages for healthcare providers to guide people living with HIV about selection of appropriate treatment partners. We will recruit 200 PLWH with treatment partners in Botswana, selected such that half show virologic failure and half show virologic suppression. We will conduct social network assessments of PLWH and their treatment partners and gather qualitative data describing their social networks in order to examine the extent to which individual, dyadic, and social network factors related to treatment partners differ between groups. We will present the results to members of the Committee for the Clinical Care of TB and HIV/AIDS in Botswana, which develops the National HIV and AIDS Treatment Guidelines for the Botswana Department of HIV/AIDS Prevention and Care in the Botswana Ministry of Health. The proposed research presents a unique opportunity to examine ways to improve use of ART in practice, a topic of critical importance to the field.